A Versatile Strategy for Characterization and ... - ACS Publications

Subscriber access provided by Kaohsiung Medical University

Article

A Versatile Strategy for Characterization and Imaging of Drip Flow Microbial Biofilms Bin Li, Sage J. B. Dunham, Joseph F. Ellis, Justin D. Lange, Justin R. Smith, Ning Yang, Travis L. King, Kensey R. Amaya, Clint Matthew Arnett, and Jonathan V. Sweedler Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.8b00560 • Publication Date (Web): 03 May 2018 Downloaded from http://pubs.acs.org on May 5, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 10 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

A Versatile Strategy for Characterization and Imaging of Drip Flow Microbial Biofilms Bin Li1,‡, Sage J. B. Dunham1,‡, Joseph F. Ellis1,‡, Justin D. Lange2, Justin R. Smith2, Ning Yang1, Travis L. King2, Kensey R. Amaya2, Clint M. Arnett2,*, Jonathan V. Sweedler1,* 1

Department of Chemistry and Beckman Institute for Advanced Science and Technology, University of Illinois at Urbana-Champaign, Urbana, IL, USA 2 Engineer Research and Development Center-Construction Engineering Research Laboratory (ERDC-CERL), Champaign, IL, USA

ABSTRACT: The inherent architectural and chemical complexities of microbial biofilms mask our understanding of how these communities form, survive, propagate, and influence their surrounding environment. Here we describe a simple and versatile workflow for the cultivation and characterization of model flow-cell based microbial ecosystems. A customized low shear, drip flow reactor was designed and employed to cultivate single and co-culture flow-cell biofilms at the air-liquid interface of several metal surfaces. Pseudomonas putida F1 and Shewanella oneidensis MR-1 were selected as model organisms for this study. The utility and versatility of this platform was demonstrated via the application of several chemical and morphological imaging techniques – including matrix-assisted laser desorption/ionization mass spectrometry imaging, secondary ion mass spectrometry imaging, and scanning electron microscopy – and through the examination of model systems grown on iron substrates of varying compositions. Implementation of these techniques in combination with tandem mass spectrometry and a two-step imaging principal component analysis strategy resulted in the identification and characterization of 23 lipids and three oligosaccharides in P. putida F1 biofilms, the discovery of interaction-specific analytes, and the observation of several variations in cell and substrate morphology present during microbially influenced corrosion. The presented workflow is well suited for examination of both single and multispecies drip flow biofilms and offers a platform for fundamental inquiries into biofilm formation, microbe-microbe interactions, and microbially influenced corrosion.

Biofilms are communities of microorganisms encased in a self-produced matrix of extracellular DNA, polysaccharides, proteins, and other biomolecules – collectively referred to as an extracellular polymeric substance (EPS).1 An EPS acts as a barrier protecting the encased organisms from environmental and biological hazards, and as a result of the biofilm phenotype, microbial communities can survive in some of the harshest environments imaginable, including on the inside of oil pipelines and at the liquid-air interface in hydrocarbon fuel tanks.2 The arrangement of microorganisms within a biofilm, and consequently the molecules that these microorganisms produce, can be highly heterogeneous and depends on many factors, including the properties of the surface, the proximity to hazardous chemicals or other microorganisms, and the local nutritional and environmental conditions.1 Biofilms play an important role in many industrial, clinical, and environmental processes. Therefore, the chemical and physical conditions underlying biofilm formation and persistence have been intensely studied by both academic and industrial researchers.3-5 For instance, microbially influenced corrosion (MIC) is a phenomenon that threatens the integrity of the world’s infrastructure – from oil and sewage pipelines, to historic buildings and military assets. A 2002 study by the United States Federal Highway Administration estimated the annual direct cost of corrosion (in 2002 dollars) to be $275 billion a year, or 3% of US GDP.6 It is estimated that the oil and gas

industry alone spends more than $8 billion each year on corrosion repair and prevention.7 Much of this corrosion is directly caused by, or greatly accelerated by, microbial biofilms.5 Because many biofilm phenomena are inherently surface specific, research into both their spatial and chemical dimensions is necessary to understand how they occur and the direct consequences of their presence. Mass spectrometry imaging (MSI) has emerged over the past two decades as an indispensable tool for coupling chemical information with two- and three-dimensional spatial information.4,8-10 In contrast to many other chemical imaging techniques, MSI can be applied to image multiple molecular species in the same experiment and can be used to study molecules of unknown composition. Although MSI is frequently employed in the study of microbial biofilms, several challenges remain in the acquisition of reliable ion images due to the hydrated, absorbent, deformable, soft, and non-uniform nature of the biofilm surface.11 Our lab has developed a number of MSI-based analytical approaches for biofilm research, including those that employ matrix-assisted laser desorption/ionization (MALDI) mass spectrometry (MS) and/or custom Buckminsterfullerene (C60+) time-of-flight (ToF) secondary ion mass spectrometry (SIMS).12-19 Here we focus on the development of a cultivation and sample preparation strategy for untargeted MS-based chemical imaging of flow-cell biofilms formed on metal substrates. Two

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

model bacteria, Pseudomonas putida strain F1 and Shewanella oneidensis strain MR-1, were selected for the development and validation of these protocols. Gram-negative bacteria belonging to the genus Pseudomonas are well-studied biofilmforming microorganisms known to promote MIC.20,21 S. oneidensis MR-1 is a member of the gamma subdivision of the Proteobacteria and is a facultative anaerobe that can respire by using oxygen or ferric iron as its terminal electron acceptor.22 Bacteria belonging to the genera are known exoelectrogens that are capable of transferring electrons through their cell membranes to external environments.23,24 The ability to shuttle electrons extracellularly has made them the focus of several MIC studies as well as microbial fuel cell research.25,26 Additionally, P. putida OUS82 and S. oneidensis MR-1 are known to exhibit both competitive and cooperative tendencies with each other in the planktonic state; however, little is known about the interactions of these two organisms during biofilm development.27 Our workflow for constructing, cultivating, and characterizing drip flow biofilms is shown in Figure 1. Briefly, the microbial communities are cultivated in a low shear drip flow reactor (DFR), flash frozen and dehydrated, imaged with SIMS, and either coated with an organic, light-absorbing matrix and imaged via MALDI MSI, or examined with scanning electron microscopy (SEM) and energy-dispersive X-ray spectroscopy (EDS). This workflow is simple and easily adaptable for studying diverse microbial biofilms using a host of analytical techniques. We demonstrate the utility and versatility of the approach by examining three culture conditions: a single culture of P. putida on stainless steel, a co-culture of P. putida and S. oneidensis on stainless steel, and a co-culture of P. putida and S. oneidensis on low-carbon steel. These samples

Page 2 of 10

were chosen for the purpose of simplified molecular discovery, and to demonstrate the capacity for examination of strainstrain chemical interactions as well as the chemical processes underlying MIC. EXPERIMENTAL Chemicals and Reagents The MALDI matrices and chemicals and reagents (of the highest purity available) were purchased from MilliporeSigma (St. Louis, MO) unless otherwise specified. Model Strains and Cultivation A custom DFR (see Supporting Information) was fabricated to accommodate the stainless steel coupons or glass slides that fit Bruker MALDI adapter plates (Bruker, Billerica, MA) and a lab-built C60-Quadrupole ToF-SIMS mass spectrometer. Substrates of 304 stainless steel and 1018 low-carbon steel were precision machined at the University of Illinois School of Chemical Sciences machine shop (see Supporting Information). The surfaces of the stainless steel substrates were conditioned with sandpaper or by vapor blasting to promote adhesion (Figure 1a and Figure S1). The Pseudomonas putida strain F1 (ATCC® 700007™) and Shewanella oneidensis strain MR-1 (ATCC® 700550™) were obtained from American Type Culture Collection (Manassas, VA). Stock cultures used for biofilm experiments were propagated overnight in nutrient broth and archived in 10% glycerol at –80˚C until use. To establish single strain biofilms on the stainless steel surfaces, bacteria were cultivated statically at 30˚C for 6 h in a logarithmic growth phase with a drip angle of 0˚. The inoculum

Figure 1. Workflow for cultivating, preserving, and characterizing drip flow biofilms. (a) Metal substrate conditioning for control of biofilm adherence; (b) cultivation of bacterial biofilms in a drip flow reactor; (c) pretreatment of biofilms via flash freezing and vacuum desiccation; (d) chemical imaging with SIMS and MALDI as well as bulk extract preparation; (e) multivariate image analysis to distinguish ions specific to different biofilm regions, such as the zone of interaction; and (f) examination of biofilm architecture and elemental composition with SEM and EDS.


Page 3 of 10 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


consisted of 5 mL 10% glycerol stock and 45 mL of 3 g/L tryptic soy broth (TSB). After 6 h, the drip angle was increased to 5˚ and a continuous flow of nutrients (TSB, 270 mg/L) was supplied at 50 mL/h via peristaltic pump through a glass flow break. Side-by-side drip flow experiments with co-cultures were performed by creating a hydrophobic barrier down the center of the coupons using a hydrophobic pen (PAP Pen Super-Liquid Blocker, Cosmo Bio USA, Inc. Carlsbad, CA). Bacteria mixture or media-only (for control experiments) was dripped on either side of the coupon and allowed to join only at the bottom third of the substrate to form a defined interaction region. To facilitate bacterial attachment, 3 g/L TSB was inoculated with 1% of each species individually. The cultures were stirred at 22 ˚C and simultaneously dripped at 50 mL/h on the coupons at a 5˚ drip angle. After 6 h of drip flow, the inoculation was halted and the reactors were fed a continuous flow of 270 mg/L TSB. The DFR was operated at 30 ˚C for 1 to 6 d under low fluid shear conditions to promote growth near the air-liquid interface. The substrates were removed at predetermined timepoints and the biofilms preserved for subsequent analysis by placing them on a steel block half submerged in liquid nitrogen (substrates did not contact the liquid nitrogen). After 30 min, the coupons were removed from the block and dried in a vacuum desiccator for a minimum of 24 h at room temperature. C60+-SIMS Imaging All SIMS measurements were conducted with a modified QSTAR XL (AB SCIEX, Framingham, MA) imaging mass spectrometer equipped with a 20 keV Buckminsterfullerene (C60+) ion beam (Ionoptika Ltd, Hampshire, UK). A detailed description of this instrument can be found in our previous work.28 Images were acquired in raster mode with a pixel size of 300 μm × 300 μm for single strain biofilms (approximate primary ion dose of 1 × 1014 ions/cm2) and 500 μm × 500 μm for co-cultures and MIC samples (approximate primary ion dose of 2 × 1013 ions/cm2). The spectrometer was set to collect secondary ions from m/z 60–850 with a Q1 bias of 15%, 25%, and 60% at m/z 100, 200, and 400, respectively. Mass calibration was performed using indium clusters (In1-7+), and imaging data was converted from wiff to imzML format with ProteoWizard29 and imzMLconverter30 with subsequent visualization and analysis in MSiReader.31 Tandem mass spectra were acquired in positive product ion mode with argon collision gas and 10–40 eV collision-induced energy as needed. MALDI Matrix Application Either 2,5-dihydroxybenzoic acid (DHB) or α-cyano-4hydroxycinnamic acid (CHCA) was applied via sublimation, using an apparatus and procedure described previously (see Supporting Information).28,32 Automated spray coating via a previously optimized protocol13 and manual airbrush application were also explored for initial experiments but both application methods resulted in disruption of the thin biofilm, most likely due to the high gas pressures involved. Therefore, sublimation was selected for subsequent experiments. The thickness of the DHB or CHCA matrix was optimized for coverage homogeneity and sensitivity. In agreement with previous experiments with nervous system tissue,28 the optimal DHB coating for MALDI imaging was found to be between 0.10 and 0.30 mg/cm2, while a matrix coating thicker than 0.40 mg/cm2

resulted in an elevated baseline and reduced sensitivity. In general, DHB was found to give better signal than CHCA and was therefore the primary matrix used throughout these experiments (Figure S2). MALDI MS Imaging Most of the MALDI MSI measurements were performed with an ultrafleXtreme MALDI-ToF/ToF mass spectrometer (Bruker), equipped with a frequency tripled Nd:YAG solidstate laser (λ = 355 nm). The laser footprint was set to “Ultra” to achieve an approximate spot diameter of 100 μm. Mass calibration was performed using DHB clusters and the Peptide Calibration Standard Kit II (Bruker). Data was acquired in positive reflection mode with pulsed ion extraction and a mass range of 100–2500 Da. The step size was set to 400 μm for single strain biofilms and 500 μm for co-cultures. Images were collected with 500 laser shots/pixel at a frequency of 1000 Hz. In situ tandem MS was conducted in LIFT mode. The MALDI imaging and initial image analysis were performed using FlexImaging 4 (Bruker). The MALDI-ToF ion images were normalized to the total ion count (TIC). A select co-culture biofilm sample was subjected to follow-up analysis with a 7T SolariX XR electrospray ionization (ESI)/MALDI Fourier transform ion cyclotron resonance (FTICR) mass spectrometer (Bruker). Coordinates from the previous ultrafleXtreme image were offset both laterally and horizontally by 250 μm to prevent resampling. The laser footprint, step size, shots, and frequency were adjusted to match the ultrafleXtreme parameters. The instrument was calibrated using sodium trifluoroacetic acid clusters and analysis was conducted in the mass range of 250–2500 Da. Principal Component Analysis of MALDI FT-ICR and SIMS Imaging Data Principal component analysis (PCA) was performed on the coculture biofilm MSI datasets. MALDI data were exported to imzML via FlexImaging. The MSI datacube was imported into MATLAB (MathWorks, Inc., Natick, MA) using a custom MATLAB script implementing imzMLConverter and MSiReader, and binned at ± 0.05 Da. A custom MATLAB script incorporating the Statistics and Machine Learning Toolbox was used to mean-center the binned datacube via singular value decomposition as well as to perform the subsequent PCA. The resulting coefficient matrix was reconstructed to produce false-colored images relating variance and spatial information, where red and cyan correspond to positive and negative variance, respectively. SEM and EDS Dehydrated biofilms were coated with approximately 7 nm of Au/Pd using a Desk II Turbo Sputter Coater (Denton Vacuum, Morristown, NJ), operated for 70 s at 64 mTorr Ar and 40 mA. SEM images and EDS spectra were acquired at an accelerating voltage of 5 kV and 20 kV, respectively, with a working distance of 10 mm with a Philips XL30 ESEM-FEG (SEM only) and a Quanta 650 ESEM-FEG (FEI Company, Amsterdam, Netherlands) under high vacuum conditions (~1 μTorr) with an Everhart-Thornley secondary electron detector. EDS elemental mapping and clustering were performed using the TEAMTM EDS Analysis System (AMETEK, Inc., Berwyn, PA) and Smart Quant mode with eZAF relative quantification.



RESULTS AND DISCUSSION Optimization of Parameters for the Preparation of Flowcell Biofilms Wide application scope, ease of operation, and reliable performance were the main aims in the design of this biofilm production approach. The DFR is often used for biofilm cultivation because it produces pronounced and highly reproducible biofilms within relatively short incubation periods.33 Biomass accumulation occurs rapidly because nutrients are supplied continually and waste is removed through gravitational flow. Additionally, the customization of DFRs according to user need can be achieved with relative ease. A custom DFR was fabricated with inspiration from a similar design by BioSurface Technologies from Bozeman, MT (DFR 110-4). The modified design was made to accommodate the specific needs of our experiments, which included the use of abnormally sized substrates and the ability to simultaneously culture multiple bacterial species on the same substrate (Figure 2a). Initial tests were performed with indium tin oxide (ITO)coated glass microscope slides (a common substrate for many vacuum MALDI measurements), as well as with both polished and roughened steel. Substantial peeling and flaking of the biofilm was observed on both the ITO-glass slides and the polished steel following sample preservation (flash freezing with liquid nitrogen and vacuum desiccation). This detachment is a common problem in many microbial MSI measure-

Figure 2. Design of the modified DFR and demonstration of cultivation flexibility. (a) Schematic of the DFR. The DFR can accommodate both single and multi-culture biofilms, and is designed to contain a defined interaction region (location represented in dashed red lines). (b) The modified DFR facilitates the study of various growth phases (top row on stainless steel) and substrates (bottom row, right to left: stainless steel single channel, stainless steel dual-channel co-culture, low-carbon steel dual-channel co-culture).

Page 4 of 10

ments.11 Roughening the metal surface with fine grit sand paper or by glass vapor blasting increased bacterial attachment during cultivation and prevented dehydration-induced detachment, without adversely affecting the MSI measurements (Figure S1). It should be noted that while single culture biofilms produced flaking on ITO-glass, the co-culture biofilms (P. putida and S. oneidensis) adhered well and did not peel (Figure S3). ITO glass slides should therefore not be excluded for consideration as substrates, especially in instances where optical microscopy is required. For example, previous work showed that it is possible to grow co-cultures of P. putida and S. oneidensis on ITO-glass slides, and map the localization of the two species with fluorescence in situ hybridization and epifluorescence microscopy.34 Single strain cultures of P. putida F1 were cultivated adjacent to one another on roughened stainless steel and harvested at 1, 3, and 6 days of growth (Figure 2b(i)). Under these conditions, biofilm formation was detected after 1 day of incubation, and significant biomass accumulation was observed within 3 days. After 6 days, the steel surface was covered with a dense, highly heterogeneous biofilm. Thus, the 3-day growth period was selected for subsequent MS measurements. We cultivated three types of biofilm samples: single culture biofilms on stainless steel for method development and molecular discovery experiments, co-culture biofilms on stainless steel for evaluation of strain-strain interactions (Figure 2b(i), and co-culture biofilms on low-carbon steel to evaluate MIC (Figure 2b(ii)). Mass deviations arising from variations in height, which can be problematic in many MSI applications, were found to be inconsequential in the present study due to the smooth, sheet-like biofilms formed in the DFR, as revealed by the optical images in Figure 2b and the SEM in Figure S4. SIMS and MALDI Imaging of P. putida F1 Biofilms Reveal a Diverse Collection of Lipids and Oligosaccharides P. putida F1 was selected as a model organism for initial experiments due to its tendency to rapidly form robust biofilms on a wide variety of surfaces. Pseudomonas is a commonly used model organism for the study of biofilm formation, with relevance in both industrial and clinical settings.35 Lipids act as useful clues for correlating the composition of a microbial community with its physiological state.36 Modulations in lipid composition help bacteria maintain adequate ultrastructure and cell membrane barriers under extreme conditions.37 A major limitation in the visualization of biofilm lipid distribution is the lack of chemical dyes and labeling techniques with applicability to individual molecules. MSI is particularly useful for capturing the spatial distribution of lipids across various mammalian and plant tissues.38 Here, a number of lipids were detected, identified, and mapped on P. putida F1 biofilms using SIMS (Figure 3a, b) and MALDI MSI (Figure 3c, d). As can be seen in the representative C60-SIMS spectrum in Figure 3a, multiple analytes were detected with SIMS, including the phosphatidylethanolamine (PE) headgroup at m/z 136.9, five diacylglycerols (DGs), and eight intact PE lipids. The identities of major ions were confirmed by in situ C60SIMS tandem MS (Figure S5-i and Table S1). Representative C60-SIMS ion images for the PE headgroup, PE(32:1), DG(32:1), DG(33:1), and DG(34:1) are provided in Figure 3b. In most cases, the strongest lipid signals were observed near the region of the initial drip flow as well as the biofilm boundary. In contrast to the molecular ions, the PE headgroup shows a more homogeneous distribution, suggesting that there


Page 5 of 10 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


may be undiscovered PE lipids spread throughout the biofilm. Quorum sensing molecules (e.g., quinolones and N-acylhomoserine lactones), which have been reported in some P. putida strains, such as KT244039 and IsoF,40 were not observed in this work. Our previous examinations of P. aeruginosa biofilms revealed that C60-SIMS is highly sensitive for quinolones and less sensitive for homoserine lactones.18,19 Two distinct ion groupings were readily detected by MALDI in the mass ranges of m/z 500–800 and m/z 1000–2000 (Figure 3c and 3e, respectively) using the optimized DHB sublimation protocol. Identification of select P. putida biofilm li-

pids was achieved through a combination of in situ MALDIToF tandem MS and ex situ high mass resolution ESI FT-ICR MS and tandem MS (see the Supporting Information). Overall, 18 lipids were detected directly on the biofilms, and 24 lipids were found in the biofilm extracts (Figures S6–S8 and Table S1), including DGs, PEs, lyso-PEs, and phosphatidylglycerols (PGs). Most identified lipids were long chain and unsaturated. Interesting variations between our data and previous reports were observed, including the observation of a number of DGs in P. putida F1 biofilms.41-43 DGs can be formed transiently as intermediates in the biosynthesis of glycerophospholipids, are known to act as secondary messengers in mammalian and

Figure 3. Representative SIMS and MALDI-ToF spectra and ion images of a P. putida 3-day single culture. (a) C60-SIMS spectrum with ions primarily arising from DG and PE lipids. The inset is an example in situ SIMS tandem MS spectrum assigned to DG(32:1). (b) C60-SIMS ion images for several intact DG and PE lipids and the PE headgroup. (c) MALDI-ToF spectrum from a region containing a high lipid abundance. The inset tandem MS spectrum is assigned to PE(32:1). (d) Representative MALDI-ToF images of two DG and three PE lipids. (e) MALDI-ToF spectrum from a region containing a high abundance of oligosaccharides. The inset tandem MS spectrum is assigned to undecaose. (f) Representative ion images showing the distributions of several oligosaccharides. Stereochemistry is not defined and scale bars are 1 cm.



microbial cells, and are implicated in the regulation of various cellular functions such as cell growth, differentiation, and apoptosis.44,45 Single species biofilms have enormous spatially segregated phenotypical variation, and accordingly, each cell can have a distinct role in promoting the fitness and propagation of the biofilm.46 The combined SIMS and MALDI analysis shown in Figure 3a–d reveals that DGs are most abundant on the outer edge of the biofilm, while PE, PG, and lyso-PE species are found throughout the biofilm, including the edge. In batch cultures, higher molecular weight membrane lipids have been shown to decrease over time as nutrients become limited.46 The relatively even distribution of high molecular weight lipids over the biofilm suggested the biota of the community is in good health, which is expected for a functioning DFR that continually delivers fresh nutrients and removes waste. Although the reasons for the distribution patterns of various lipids have yet to be established, the differences in distribution may be due to the microbial community’s response to local physical and biological factors, including the shear force of the drip flow, local cell density, nutritional conditions, and oxygen content. The second cluster of ions in the mass range of m/z 1000–2000 (Figure 3e) was found to have an ion-to-ion mass difference of 162 Da. Subsequent LIFT-ToF/ToF analysis revealed these ions to be oligosaccharides (Figure 3e and Figure S9(i)). Due to the complex structural arrangement of many oligosaccharides, the assignments made here are tentative and the stereochemistry is undefined. As shown in Figure 3f, ions associated with oligosaccharides are preferentially found at the biofilm boundary, with some increased intensity near the center. No oligosaccharide-associated ions were observed from either the TSB media or the nutrient broth (Figure S9(ii)), and the oligosaccharide ions were only detected directly from the P. putida biofilm (Figure S9(iii)). To our knowledge, this is the first MSI observation of oligosaccharides in P. putida biofilms. While we did not directly investigate the roles of these compounds, our results suggest that oligosaccharides may be closely involved in the formation and maturation of bacterial biofilms. Additionally, these data demonstrate that MALDI imaging is amenable for in situ detection and localization of oligosaccharides in biofilms. PCA-assisted MSI of P. putida and S. oneidensis Cocultures Reveals Molecules Specific to Each Strain and the Region of Interaction Manual inspection of MSI datasets is both computationally and intellectually cumbersome, particularly for large and complex samples like those examined here. Fortunately, there have been significant advances in the application of multivariate analysis to MSI data. One of the most common multivariate methods used in MSI is imaging PCA, where variables that exhibit co-variance are grouped together into a lower dimension space, allowing for visualization of multivariate data with minimal loss of information.47 In general, two types of imaging PCA are used: targeted PCA, which accounts for the variance of a previously selected small group of analytes in the dataset, and untargeted PCA, which includes every variable in the dataset and is especially useful for analyte discovery. In this analysis, we applied a two-step combined approach. First, an initial untargeted PCA is used to discriminate ions that arise from the chemical background (e.g., culture medium

Page 6 of 10

Figure 4. Distinct ion distributions in co-culture (a, b) P. putida (left channel) and (a, b) S. oneidensis (right channel) biofilms, and their (c, d) interaction region (bottom center) revealed through two-step (a, b) MALDI and (c, d) SIMS imaging PCA. (a) MALDI PCA results showing ions independently associated with P. putida (positive loadings) and S. oneidensis (negative loadings); (b) MALDI ion images showing the distribution of the most significant contributors. (c) Two-step SIMS imaging PCA showing ions specific to the interaction zone on the positive loadings; and (d) SIMS ion images of the most prominent ions in the interaction region. Scale bars represent 1 cm in a–d. See the main text for the identity of the molecular species.

and DHB matrix) from ions specific to the biofilm. In general, the first principal component (PC1) discriminates on this basis, with the positive loadings comprised of the chemical background and the negative loadings comprised of biofilmspecific compounds (Figure S10). A second round of targeted PCA is performed only on the ions contributing to the negative loadings of PC1, enabling determination of covariance based on the specific subpopulation of ions relevant to the bacteria under study. For this multistep analysis, PC1 of the second step (targeted PCA) produces empty negative loadings (resulting from the zero values in the matrix) and therefore, only lower variance PCs contain relevant information. Applying this two-step MALDI imaging PCA to a co-culture biofilm containing P. putida (Figure 4a, left ion image) and S. oneidensis (Figure 4a, right ion image) reveals compounds specific to each bacterial species. The positive loadings of PC2 correspond to ions most abundant in P. putida biofilms, and


Page 7 of 10 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


the negative loadings correspond to ions most abundant in S. oneidensis biofilms (Figure 4a, right ion image). Several ions found to be distinct to the two species were subjected to identification via tandem MS. S. oneidensis-specific lipids included DG(29:2) (m/z 523.473) and PE(30:0) (m/z 702.447). P. putida-specific lipids included PE(32:1) (m/z 728.462) and DG(33:1) (m/z 563.504). Ion images for the most significant distinguishing ions are shown in Figure 4b. Confirmatory controls were performed for both P. putida and S. oneidensis using single-strain DFR biofilm on the dual-channel system. For each species, these ions are observed to be specific for each class (Figure S11). Remarkably, two-step PCA of the C60-SIMS data pulled out ions pertaining to both microbial strains as well as the interaction region (Figure 4c, d). Ions specific to, or more abundant in the interaction region included m/z 120.1, 268.2, 417.2, and 523.5. Single strain dual-channel controls showed that these ions are unique to or at a higher abundance in the interaction region between P. putida and S. oneidensis (Figure S11). Tandem MS was attempted for the three most-prominent ions (Figure S5(ii)) and each exhibited fragmentation patterns similar to DG lipids. However, conclusive identifications were unobtainable. Microbe-microbe chemical interactions are complex, so that elucidation of meaningful information often requires the development of new and innovative methodologies. The results provided in Figure 4 demonstrate that our two-channel DFR and the associated analytical methodologies are especially good for elucidating molecular information in spatially defined sample regions. Additional MSI techniques could be readily incorporated into the current workflow, such as laser ablation inductively coupled plasma (LA-ICP) MSI, which can be applied for the investigation of the siderophore pathways normally activated during microbe-microbe interaction.48 P. putida Swims Against Gravity and Shear Flow to Colonize Nutrient-rich Surfaces The observation of ions associated with P. putida on the S. oneidensis side of co-cultures (e.g., Figure 4) led us to ask, “Are these compounds present in both bacterial strains, or are P. putida biofilms forming on both sides of the substrate?” SEM was used to examine control samples containing a P. putida drip on one side and a nutrient-only drip on the other; we found that P. putida actually migrates up the opposing channel against the nutrient flow. Electron micrographs of the P. putida channel (Figure S12) show cell agglomerations covering the surface and seeding the steel crevices. As in the coculture samples shown in Figure 4, these electron micrographs are mostly devoid of the biofilm phenotype. Remarkably, the nutrient-only channel also shows crevasses seeded with P. putida cell clusters (Figure S12). This same behavior is not observed in controls with nutrients dripped on both sides, nor in controls with S. oneidensis on one side and media on the other, demonstrating that it is not the result of reactor contamination. This suggests that the previously observed P. putida-associated ions found on the S. oneidensis side of the coupon may actually arise from climbing P. putida. We are not the only lab to observe this phenomenon of climbing bacteria. In one noteworthy example, Shireen Kotay and coworkers49 seeded sink drains with green fluorescent proteinexpressing Escherichia coli and found that the bacteria climb up the drain pipe (against the flow of the waste water) at an

average rate of 2.5 cm per day. Interestingly, examination of single culture controls revealed that S. oneidensis does not readily form biofilms under shear-force flow. Following cocolonization with P. putida, an increase in S. oneidensis biofilms was observed, as demonstrated by lipids directly associated with S. oneidensis (Figure 4). Multimodal Investigations of P. putida- and S. oneidensisInfluenced Corrosion The scientific community’s mechanistic understanding of MIC is in its early phases; however, evidence suggests that bacteria follow a general trend of colonizing a surface, forming biofilms, and initiating the secretion of chemicals capable of oxidizing or reducing iron, sulfur, and manganese.50 In most realworld cases, these microbial communities are made up of multiple species, living both competitively and synergistically.50 In this work, we set out to test the viability of our platform for studying MIC by cultivating P. putida and S. oneidensis adjacent to one another on low-carbon steel (Figure 5). Other steel

Figure 5. Characterization of MIC on low-carbon steel via SIMS and SEM. (a) Optical and (b) SIMS TIC images of a coculture system on a corroded surface. (c) SIMS spectra of regions i and iii contain predominantly high mass defect ions. (d) SEM of (i) bacteria near S. oneidensis nutrient drip, which show signs of cell lysis and degradation, (ii) bacteria in the P. putida channel showing increased size and iron oxide on the cellular wall, (iii) interaction zone bacteria that display changes in morphology, and (iv) pitted surface containing bacteria formed during corrosion. Red boxes are magnified by a factor of 10 with respect to the parent image. Scale bars represent 1 cm (a and b), 5 µm (d(i–iii)), and 250 µm (d(iv)).



variants were tested, but were observed to develop either excessive amounts of corrosion, or no corrosion after three days of biofilm growth. Figure 5a shows an optical image of a representative sample, with S. oneidensis on the left and P. putida on the right, and significant corrosion on both sides. The sample was first examined with SIMS (Figure 5b, c), then coated with a thin layer of Au/Pd and examined with SEM (Figure 5d). The SIMS TIC image (Figure 5b) shows that most ions arise from the area surrounding the MIC, with a region of high abundance near the top of the P. putida side. Surprisingly, most of the ions observed in the SIMS spectra were distinct from those observed in the analogous culture on stainless steel (Figure 4). Our attempts at identifying the highest intensity ions were unsuccessful, but the high mass defect suggests an inorganic composition (Figure 5c). MALDI MSI was also performed on a separate MIC coculture sample and several ions specific to either P. putida or S. oneidensis were observed (Figure S13). Unfortunately, under the sample preparation steps used here, the ion signal was too low for additional conclusions. We hypothesize that inorganic products resulting from MIC of steel hinder ionization of lipids and other organic species in positive ion mode. Although the chemical information garnered from these investigations was unsatisfactory, the large abundance of ions observed via SIMS suggests that further optimization of the sample preparation procedure and additional ion identification efforts will lead to results of use to industries affected by MIC. We next investigated specific regions of the sample using a combination of SEM and EDS (Figure 5d, Figure S14, and Figure S15). The SEM images in Figure 5d(i–iv) correspond to the analogously labeled regions in Figure 5a. In contrast to the co-culture on stainless steel (Figure 4), bacteria on lowcarbon steel do not appear to form robust biofilms. The cells formed clumps and chains as opposed to the previously observed ordered cellular arrays encased in EPS. Within the corroded section of the S. oneidensis side of the sample (Figure 5d(i)), the long pill-shaped cells appear to be damaged or partially lysed, with small holes present over the cell bodies. Many cells are also coated in small (< 100 nm) bulbous structures, which may indicate cell lysis. It is unlikely that these morphological characteristics arose from vacuum exposure, as other cells in the immediate vicinity are intact. Bacteria on the P. putida side (Figure 5d(ii)) appear to be healthier (i.e., not ruptured); however, upon closer inspection, a filamentous phenotype is observed. The expression of a filamentous phenotype has been previously observed in P. putida during environmental stressors,51 and is seen with S. oneidensis MR-1 during biofilm formation and in the presence of iron.52,53 Near the interaction zone (Figure 5d(iii)) the cells appear more similar in size and shape to those observed in the P. putida region, and a significant quantity of iron oxide is found to be uniformly surrounding the cell with an exquisite nanostructure (EDS of representative cells is shown in Figure S15). Adhesion of iron and iron nanoparticles to P. putida and S. oneidensis has been demonstrated,54-56 but this extent of iron oxide uniformly coating the surface of the bacteria has not been previously reported. Near the bottom of the S. oneidensis side of the sample (Figure 5(iv)), bacteria are packed into the corroded sections and large structures are present. Inspection with EDS revealed that these structures are primarily composed of iron oxide (Figure S15).

Page 8 of 10

CONCLUSIONS In this work, we developed a modified DFR and applied it to a series of chemical and morphological studies of biofilms on steel substrates. Following optimization of the substrates, growth modes, and sample preparation methodologies, we demonstrate the use of the DFR to study lipid and oligosaccharide distribution across a single culture biofilm using both MALDI and SIMS imaging. Using a combination of in situ tandem MS and FT-ICR MS-based accurate mass characterization of extracts, we elucidate the partial identities of lipids and oligosaccharides expressed by biofilm-bound P. putida and map their distribution across the colony surface. We studied a co-culture system comprised of P. putida and S. oneidensis on stainless steel using a combination of SIMS and MALDI imaging followed by two-step imaging PCA. This study revealed ions exclusively associated with each species, as well as ions found to be elevated in the interaction zone. Next, we examined an MIC sample comprised of a low-carbon steel coupon with a co-culture of P. putida and S. oneidensis. SIMS imaging revealed several ions that could only be found during MIC. Follow-up SEM analysis showed clumps of bacteria residing in crevasses, many of which were expressing the filamentous phenotype. Remarkably, most of the growth appeared to be occurring without the presence of a strong biofilm phenotype, and many of the cells were coated in iron oxide. This work establishes (among other avenues) a platform for studying biofilm metabolites and biomarkers, the interaction of multiple microbial species, and the ability to alter the environment for the study of MIC. The coupling of MSI with SEM and EDS provides a valuable opportunity to correlate unbiased and untargeted chemical information with biofilm morphology and microbial distribution. MSI has evolved into a useful technique for chemical imaging in microbiology, and has been used throughout these studies. Of course, other imaging modalities, such as fluorescence in situ hybridization and Raman spectroscopy, can be incorporated into our workflow for additional chemical information. The adaptable DFR and associated methods described here can be applied and extended to monitor the spatially adaptive response of bacterial biofilms under a wide variety of experimental conditions.

ASSOCIATED CONTENT Supporting Information The Supporting Information includes Additional Methods (Drip Flow Reactor (DFR) Design and Construction, Customization of Metal Substrates, Sublimation of DHB and CHCA Matrix, Lipid Extraction and Analysis by MALDI-ToF/ToF, and HighResolution FT-ICR MS); Figures S1–S15, and Table S1 as noted in the text. The Supporting Information is available free of charge on the ACS Publications website.

AUTHOR INFORMATION Corresponding Authors Jonathan V. Sweedler Postal: 600 S. Mathews Ave, Urbana, IL, 61801, USA Email: [email protected] Tel: +1 217-244-7359 Fax: +1 217-244-8068 Clint M. Arnett Postal: 2902 Newmark Drive, Champaign, IL 61822, USA


Page 9 of 10 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Email: [email protected] Tel: +1 217-398-5507 Fax: +1217-373-7266

Author Contributions The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. ‡These authors contributed equally.

KEYWORDS Mass Spectrometry Imaging, Secondary Ion Mass Spectrometry Imaging, Energy-dispersive X-ray Spectroscopy, Scanning Electron Microscopy, Microbial Influenced Corrosion, Biofilms, Pseudomonas putida, Shewanella oneidensis, Microbe-Microbe Interactions, Multimodal Imaging

ACKNOWLEDGMENTS The authors would like to acknowledge intellectual contributions from Troy Comi and Elizabeth Neumann. SEM support provided by Catherine Wallace and Scott Robinson from the Beckman Institute Imaging Technology Group, which is partially supported by the National Science Foundation Award No. DBI-9871103. Funding for this study was provided through a cooperative agreement with the Engineer Research and Development CenterConstruction Engineering Research Laboratory under agreement number W9132T-15-2-0006. JFE and SJBD gratefully acknowledge support from the Springborn Endowment, and SJBD received partial financial support through the NSF-GRFP. Although this research was sponsored by the Department of Defense, the content of the information does not necessary reflect the position or policy of the government and no official endorsement should be inferred.

REFERENCES (1) Donlan, R. M. Emerging Infect. Dis. 2002, 8, 881-890. (2) Flemming, H. C.; Wingender, J. Nat. Rev. Microbiol. 2010, 8, 623-633. (3) Hall-Stoodley, L.; Costerton, J. W.; Stoodley, P. Nat. Rev. Microbiol. 2004, 2, 95-108. (4) Dunham, S. J.; Ellis, J. F.; Li, B.; Sweedler, J. V. Acc. Chem. Res. 2017, 50, 96-104. (5) Beech, W. B.; Sunner, J. Curr. Opin. Biotechnol. 2004, 15, 181-186. (6) Koch, G. H.; Brongers, M.; Thompson, M.; Virmani, Y. P.; Payer, J., Corrosion Cost and Preventative Strategies in the United States. Federal Highway Administration; Report# FHWA-RD-011562002 (7) Bartling, C. Pipeline & Gas Journal 2016, 243. (8) Cornett, D. S.; Reyzer, M. L.; Chaurand, P.; Caprioli, R. M. Nat. Methods 2007, 4, 828-833. (9) Debois, D.; Jourdan, E.; Smargiasso, N.; Thonart, P.; De Pauw, E.; Ongena, M. Anal. Chem. 2014, 86, 4431-4438. (10) Watrous, J. D.; Dorrestein, P. C. Nat. Rev. Microbiol. 2011, 9, 683-694. (11) Yang, J. Y.; Phelan, V. V.; Simkovsky, R.; Watrous, J. D.; Trial, R. M.; Fleming, T. C.; Wenter, R.; Moore, B. S.; Golden, S. S.; Pogliano, K.; Dorrestein, P. C. J. Bacteriol. 2012, 194, 60236028. (12) Masyuko, R. N.; Lanni, E. J.; Driscoll, C. M.; Shrout, J. D.; Sweedler, J. V.; Bohn, P. W. Analyst 2014, 139, 5700-5708. (13) Li, B.; Comi, T. J.; Si, T.; Dunham, S. J.; Sweedler, J. V. J. Mass Spectrom. 2016, 51, 1030-1035. (14) Si, T.; Li, B.; Zhang, K.; Xu, Y.; Zhao, H.; Sweedler, J. V. J. Proteome Res. 2016, 15, 1955-1962.

(15) Si, T.; Li, B.; Comi, T. J.; Wu, Y. W.; Hu, P. F.; Wu, Y. Y.; Min, Y. H.; Mitchell, D. A.; Zhao, H. M.; Sweedler, J. V. J. Am. Chem. Soc. 2017, 139, 12466-12473. (16) Lanni, E. J.; Masyuko, R. N.; Driscoll, C. M.; Aerts, J. T.; Shrout, J. D.; Bohn, P. W.; Sweedler, J. V. Anal. Chem. 2014, 86, 9139-9145. (17) Lanni, E. J.; Masyuko, R. N.; Driscoll, C. M.; Dunham, S. J. B.; Shrout, J. D.; Bohn, P. W.; Sweedler, J. V. Anal. Chem. 2014, 86, 10885-10891. (18) Baig, N. F.; Dunham, S. J.; Morales-Soto, N.; Shrout, J. D.; Sweedler, J. V.; Bohn, P. W. Analyst 2015, 140, 6544-6552. (19) Dunham, S. J.; Comi, T. J.; Ko, K.; Li, B.; Baig, N. F.; Morales-Soto, N.; Shrout, J. D.; Bohn, P. W.; Sweedler, J. V. Biointerphases 2016, 11, 02A325. (20) Wang, G.; Chai, K.; Wu, J. Y.; Liu, F. C. Int. Biodeterior. Biodegrad. 2016, 115, 156-163. (21) Li, H. B.; Zhou, E. Z.; Zhang, D. W.; Xu, D. K.; Xia, J.; Yang, C. G.; Feng, H.; Jiang, Z. H.; Li, X. G.; Gu, T. Y.; Yang, K. Sci. Rep. 2016, 6, 20190. (22) Bretschger, O.; Obraztsova, A.; Sturm, C. A.; Chang, I. S.; Gorby, Y. A.; Reed, S. B.; Culley, D. E.; Reardon, C. L.; Barua, S.; Romine, M. F.; Zhou, J.; Beliaev, A. S.; Bouhenni, R.; Saffarini, D.; Mansfeld, F.; Kim, B. H.; Fredrickson, J. K.; Nealson, K. H. Appl. Environ. Microbiol. 2007, 73, 7003-7012. (23) Logan, B. E. Nat. Rev. Microbiol. 2009, 7, 375-381. (24) El-Naggar, M. Y.; Wanger, G.; Leung, K. M.; Yuzvinsky, T. D.; Southam, G.; Yang, J.; Lau, W. M.; Nealson, K. H.; Gorby, Y. A. Proc. Natl. Acad. Sci. U. S. A. 2010, 107, 18127-18131. (25) Majumder, D.; Maity, J. P.; Tseng, M. J.; Nimje, V. R.; Chen, H. R.; Chen, C. C.; Chang, Y. F.; Yang, T. C.; Chen, C. Y. Int. J. Mol. Sci. 2014, 15, 16772-16786. (26) Kouzuma, A.; Kasai, T.; Hirose, A.; Watanabe, K. Front. Microbiol. 2015, 6, 609. (27) Wang, V. B.; Chua, S. L.; Cai, Z.; Sivakumar, K.; Zhang, Q.; Kjelleberg, S.; Cao, B.; Loo, S. C.; Yang, L. Bioresour. Technol. 2014, 155, 71-76. (28) Lanni, E. J.; Dunham, S. J.; Nemes, P.; Rubakhin, S. S.; Sweedler, J. V. J. Am. Soc. Mass Spectrom. 2014, 25, 1897-1907. (29) Chambers, M. C.; Maclean, B.; Burke, R.; Amodei, D.; Ruderman, D. L.; Neumann, S.; Gatto, L.; Fischer, B.; Pratt, B.; Egertson, J.; Hoff, K.; Kessner, D.; Tasman, N.; Shulman, N.; Frewen, B.; Baker, T. A.; Brusniak, M.-Y.; Paulse, C.; Creasy, D.; Flashner, L., et al. Nat. Biotechnol. 2012, 30, 918-920. (30) Race, A. M.; Styles, I. B.; Bunch, J. J. Proteomics 2012, 75, 5111-5112. (31) Robichaud, G.; Garrard, K. P.; Barry, J. A.; Muddiman, D. C. J. Am. Soc. Mass Spectrom. 2013, 24, 718-721. (32) Hankin, J. A.; Barkley, R. M.; Murphy, R. C. J. Am. Soc. Mass Spectrom. 2007, 18, 1646-1652. (33) Goeres, D. M.; Hamilton, M. A.; Beck, N. A.; BuckinghamMeyer, K.; Hilyard, J. D.; Loetterle, L. R.; Lorenz, L. A.; Walker, D. K.; Stewart, P. S. Nat. Protocols 2009, 4, 783-788. (34) Arnett, C. M.; Lange, J. D. T., Method of Localizing and Differentiating Bacteria within Biofilms Grown on Indium Tin Oxide: Spatial distribution of Exo-electrogenic Bacteria within Intact ITO Biofilms via FISH; ERDC/CERL TR-17-42; U.S. Army Engineer Research and Development Center: Champaign, IL, 2017, DOI: http://dx.doi.org/10.21079/11681/25701. (35) Nikel, P. I.; Martinez-Garcia, E.; de Lorenzo, V. Nat. Rev. Microbiol. 2014, 12, 368-379. (36) van Baar, B. L. FEMS Microbiol. Rev. 2000, 24, 193-219. (37) Zhang, Y. M.; Rock, C. O. Nat. Rev. Microbiol. 2008, 6, 222233. (38) Goto-Inoue, N.; Hayasaka, T.; Zaima, N.; Setou, M. Biochim. Biophys. Acta 2011, 1811, 961-969.



(39) Diggle, S. P.; Lumjiaktase, P.; Dipilato, F.; Winzer, K.; Kunakorn, M.; Barrett, D. A.; Chhabra, S. R.; Camara, M.; Williams, P. Chem. Biol. 2006, 13, 701-710. (40) Steidle, A.; Allesen-Holm, M.; Riedel, K.; Berg, G.; Givskov, M.; Molin, S.; Eberl, L. Appl. Environ. Microbiol. 2002, 68, 6371-6382. (41) Fang, J. S.; Barcelona, M. J.; Alvarez, P. J. J. Org. Geochem. 2000, 31, 881-887. (42) Pinkart, H. C.; White, D. C. J. Bacteriol. 1997, 179, 42194226. (43) Fang, J.; Barcelona, M. J.; Alvarez, P. J. Appl. Microbiol. Biotechnol. 2000, 54, 382-389. (44) Carrasco, S.; Merida, I. Trends Biochem. Sci. 2007, 32, 2736. (45) Rhome, R.; Del Poeta, M. Annu. Rev. Microbiol. 2009, 63, 119-131. (46) Benamara, H.; Rihouey, C.; Abbes, I.; Ben Mlouka, M. A.; Hardouin, J.; Jouenne, T.; Alexandre, S. PLoS One 2014, 9, e108478. (47) Jones, E. A.; Deininger, S. O.; Hogendoorn, P. C.; Deelder, A. M.; McDonnell, L. A. J. Proteomics 2012, 75, 4962-4989.

Page 10 of 10

(48) Wakeman, C. A.; Moore, J. L.; Noto, M. J.; Zhang, Y.; Singleton, M. D.; Prentice, B. M.; Gilston, B. A.; Doster, R. S.; Gaddy, J. A.; Chazin, W. J.; Caprioli, R. M.; Skaar, E. P. Nat. Commun. 2016, 7, 11951. (49) Kotay, S.; Chai, W. D.; Guilford, W.; Barry, K.; Mathers, A. J. Appl. Environ. Microbiol. 2017, 83, e03327-03316. (50) Kip, N.; van Veen, J. A. ISME J. 2015, 9, 542-551. (51) Crabbé, A.; Leroy, B.; Wattiez, R.; Aertsen, A.; Leys, N.; Cornelis, P.; Van Houdt, R. BMC Microbiol. 2012, 12, 282. (52) Godeke, J.; Paul, K.; Lassak, J.; Thormann, K. M. ISME J. 2011, 5, 613-626. (53) Binnenkade, L.; Teichmann, L.; Thormann, K. M. Appl. Environ. Microbiol. 2014, 80, 5304-5316. (54) Parikh, S. J.; Chorover, J. Langmuir 2006, 22, 8492-8500. (55) Li, B.; Logan, B. E. Colloids Surf. B. Biointerfaces 2004, 36, 81-90. (56) Lee, A. K.; Newman, D. K. Appl. Microbiol. Biotechnol. 2003, 62, 134-139.

FOR TOC ONLY


A Versatile Strategy for Characterization and ... - ACS Publications

Recommend Documents